I believe that some sell-side analysts - particularly the bearish ones who have laughably called for $16-18 price targets on shares of Intel (INTC) - are quite possibly now willing to say absolutely anything to stop the Intel rocket ship that has been furiously gaining momentum. Not only are shares up 20% YTD, but Intel has been gaining momentum in the tablet space that has managed to surprise the bears. I suggest that you do not listen to these bearish sell-side analysts, and that you do not let them scare you out of your shares. They've been wrong up until now, and they will continue to be wrong as 2013 plays out.
Uh, Everyone Could Do It, But Intel Couldn't?
It should have been a massive hint that if Qualcomm (QCOM), Freescale (FSL), MediaTek, Allwinner, Nvidia (NVDA), and Apple (AAPL) could all design and implement competent mobile SoCs, so could Intel. For a long time, there was a myth that "x86" was inherently too power hungry for tablets and phones. The problem with this is that most people investing in semiconductor stocks don't actually understand the difference between micro-architecture and instruction-set architecture, so the non technically literate genuinely believed that the Intel instruction set (which is - very loosely - just a specification of what commands that a chip can understand) somehow prohibited lower-power parts. They did not realize that it was simply all about setting out to hit the right design targets for lower power in the micro-architecture, which is the actual design that implements the execution of said instruction set.
When Intel announced its low-power roadmap in 2011, Intel bulls knew that it would only be a matter of time before Intel was fiercely competitive in this space, since Intel would be implementing smartphone/tablet optimized micro-architectures:
(click to enlarge)
Furthermore, the public perception seems to have been, "oh Intel needs to be one process node ahead to be on par with ARM (ARMH)". Of course, this too, is nonsense, unless one wants to assume that the micro-architecture and physical design folks over at Intel are completely incompetent (they're not). What a process advantage does is actually allow you to do more in a given area/power envelope than competitors assuming all else equal. At 32nm, watt-for-watt, Intel's "Clover Trail+" competes very well with parts from Qualcomm and Nvidia.
So it seems that Intel's current parts are competitive, and that the next gen parts (available this year) will probably have a significant performance/watt lead over the majority of its peers. It seems that Wall Street is finally wise to this, particularly after the "Bay Trail" demos at Computex and the "Clover Trail+" design wins at Samsung, ASUS, Acer, and others.
Well, now that Intel proved the "x86 myth" completely wrong, the bears with their $18 price targets have a new buzzword to try to spread around - "cannibalization."
Oh No! Bay Trail Will Cannibalize Haswell!
So, check this out...here's an example of the negativity after Intel finally showed its mobile teeth:
"If during the holiday season, an OEM or a consumer has to choose between buying a Silvermont Win 8.1 Tablet (~$40 INTC content) and a Haswell 8.1 NB (~$110 content), assuming some cannibalization, the net processor revenue impact for Intel is negative." - Stern Agee's Vijay Rakesh
Okay, so let me get this straight... The analysts who are saying ARM-based tablets are cannibalizing PC sales now think that because Intel has a lower power solution, that this will cannibalize... PC sales? There are a couple of problems with this analysis:
- The notebooks likely to be cannibalized by any Atom parts will be the Pentium/Celeron ones, so the cannibalization is not likely anywhere near trading ~$110 of Intel content for $40-50.
- Even if the dollar content goes down slightly at the low end, the Atom parts are much smaller and aren't disabled, much larger die, so gross margins may actually improve slightly
- Even if there is cannibalization, the Atom parts target much higher volume devices (7" to 10" tablets, cheaper notebooks/convertibles, and the whole suite of Android devices). So it would really be more along the lines of 2+ Atoms for each low end Celeron/Pentium
- Even if the above are all understated, the fear is that ARM based SoCs are cannibalizing the low end anyway, so it's better to sell an Intel chip than none at all, improving the current operating situation from what analysts had previously believed
So what exactly is the argument that are the bears trying to push? Intel dramatically expanding its TAM, defending its classical PC business vigorously by trying to make the devices more compelling, and is putting out competitive products in the new segments to gain share. In these new markets, there is incredible secular growth and since Intel also stands to gain market segment share, it is a win/win.
If you want to take profits on or even short Intel, go ahead. But you'd better have a tighter thesis than "Intel's tablet chips will be too good." At the end of the day, the computing market is exploding, so any "cannibalization" of the low end PC market will be more than offset by the explosion of volume in the tablet markets, and then for Intel during 2014, the smartphone market (which does not cannibalize PCs). The future, in my view, is brighter than ever for Intel, and I think that it's going to take a lot more than a few bearish analysts trying to justify their bearish stances to drive any meaningful downside to this still-hated stock.